from captured individuals, in addition to matching call characteristics with those reported in the 

 literature. 



At the time detector units were placed we recorded 1) an index of sky conditions from 0-6 

 (clear to showers), 2) Beaufort wind scale from 0-5 (no wind to winds 19-24 mph), and 3) 

 ambient temperature (see Appendix 1). We assumed that conditions at the time units were 

 placed provided an index of nocturnal conditions. 



Mist-nets were set singly or in pairs and run for 1 .5 to 2 hours beginning at dusk. Netting was 

 usually conducted near a water source or where bat activity would be concentrated. Nets were 

 set across creeks, beaver ponds, an open field, and a mine adit. Some mines were examined for 

 bat activity, but potential roosts in caves or free cavities were not checked, despite the presence 

 of several in the region. Mine inspection included visual examination and entry if possible, use 

 of ulfrasound bat detectors near the portals in most cases, and use of mist-nets in one case. Use 

 of detectors and nets at mines followed procedures employed at other localities. 



An attempt was made to visit the six Districts of the Kootenai National Forest (Cabinet, 

 Fisher River, Fortine, Libby, Rexford, and Three Rivers) each year, but coverage was not 

 uniform across all Districts. Survey sites were chosen based on accessibility and the presence of 

 water nearby, as bats tend to concentrate their foraging activity over water sources where insects 

 are abundant (all bats in this region are insectivorous). However, forest clearings or narrow 

 corridors within forest stands were also monitored. Habitat data were collected for each site (see 

 field form in Appendix 1), and an attempt was made to sample different cover types. 



Sample sizes from different habitats were too few to analyze separately other than in a broad- 

 spectrum manner. Habitat data were placed into three general forest-structure categories (old- 

 growth/mature, "disturbed", "other"). "Distiirbed" stands comprised all recent to pole-sapling 

 stages in seed-free cut, clearcut, and bum stands, whereas the "other" category was comprised of 

 habitats not encompassed by the first two categories (e.g., young forest, mixed conifer/deciduous 

 stands, aspen stands, etc.). The "other" category included too few sites for use in statistical 

 analyses of bat activity in forest of different structure. Each category pooled data from all forest 

 types (e.g., Douglas-fir, western redcedar, western larch, ponderosa pine, Engelmann spruce, 

 subalpine fir, grand fir), although most sites sampled were in mixed-conifer stands. Riparian 

 sites were included as a fourth category for some comparisons with forested categories, 

 particularly old-growth/mature stands. Riparian sites were located within forested areas but the 

 vegetation was dominated by typical deciduous phreatophyte genera such as cottonwood 

 (Populus), birch (Betula) and willow (Salix). The next level of analysis for future studies should 

 be of bat presence in stands of different forest structure within a forest type. Statistical analyses, 

 where used, follow standard procedures described in Sokal and Rohlf (1981), with statistical 

 significance assumed when P < 0.05. 



